Somebody just wrote the obituary for the blogging business, and it deserves about six minutes of your attention.
A digital asset investor named Daniel Stanica tracked 100 once successful blogs for four years, from April 2022 to April 2026, then published what he found. It is not comfortable reading. The median blog in his sample lost 85 percent of its organic traffic. Twelve of them now sit at zero organic visits. Generic personal finance content, the most summarizable category he measured, fell by a median of 99 percent.
Only 21 of the 100 grew.
The uncomfortable part is why
The blogs that got wiped out were not lazy operations. Most of them were competent. They ranked for years. What they had in common was the shape of their content: informational queries. How to. What is. Best ways to. Exactly the kind of question that Google, or ChatGPT, can now answer completely on the results page without ever sending a human to your website.
That is the entire mechanism. If a machine can assemble your answer from three other sources, the reader never needs to arrive. Your post can be fresh, well written and keyword perfect, and still be invisible, because the click that used to justify it simply does not happen anymore.
This is where a lot of AI content strategies quietly fail. Point a language model at “10 tips for first time homebuyers” and you get something that reads fine and is, by definition, summarizable. You have automated the production of precisely the content the answer engines were built to replace. New content, fewer clicks, and a dashboard that takes a year to tell you.
What actually survived
Stanica’s survivors shared one trait, and it is worth sitting with: firsthand experience that a machine cannot produce or summarize from somewhere else. Recipes from people who actually cooked the dish. Projects from people who actually built the thing. Local knowledge from someone who was actually standing there.
“The thing that protected you from the algorithm is the same thing that protects you from the answer engine.”
Daniel Stanica, The Great Blogging Collapse
Ask a general purpose AI why inventory is tight in a given market and it will hand you a national statistic. Ask an agent who has been writing offers in that market for fifteen years and you get something else entirely. You get the fact that their sellers are sitting on 3 percent mortgages, and that the math is keeping good sellers exactly where they are. You get the specific neighborhood where the squeeze is worst. You get the property tax reset that makes moving across town cost more than it looks like it should.
No model can summarize that from three other sources, because it was never in three other sources. It was in a person’s head, and the only way to get it out is to ask them.
This is the whole reason we put a human in the room
Every article we ship starts with a real expert talking, for about an hour, about what they actually know. That interview is the raw material. Everything downstream of it is structure, optimization and distribution wrapped around a human being’s actual expertise.
That is not a stylistic preference. It is the one input Stanica’s data says still works.
It also means we guard it. When an expert names a real person, a real place or a real number, that detail survives to publish with their name attached to it. Every pull quote in every article gets checked back against what the human actually said in the interview. The moment an editing pass smooths a named source into “a leading industry provider,” the exact thing that made the article un-summarizable is gone, and you have paid good money to publish the same slop the answer engines just finished burying.
Stop selling traffic. Start owning authority.
Here is the shift worth internalizing, and it is a hard one for anyone who has spent a decade watching a rankings report.
Search is an acquisition channel now, not the business. The goal is no longer “we will rank.” The goal is to be credible, to own an audience you can reach without asking Google for permission, and to be the source the AI cites when it answers a question on your behalf. Being summarized is fine, as long as you are the one being credited when it happens.
That is a fundamentally different product than a blog post. It also happens to be the one still standing.
If you are publishing AI content today and you are not certain which side of that line it falls on, it is worth finding out before your traffic tells you. Send us the site. We are happy to take a look.
Want a content rhythm like this built into your CRM?
Contact ReadTomato
